Unproductive Labor and the "Mage Heresy" III: Rejoinder to Fred

Murray Smith (msmith@SPARTAN.AC.BROCKU.CA)
Mon, 9 Feb 1998 16:11:38 -0500 (EST)

Many thanks to Fred for his extended post of 3 February, continuing the
discussion of "Unproductive Labor and the 'Mage Heresy'." It is an
excellent restatement and defense of what Fred calls the "standard
interpretation" of how Marx treats the costs of unproductive labor in
relation to the fundamental value categories of his system -- an
interpretation with which I continue to disagree.

Before addressing the finer points of Fred's philological analysis of the
contested quotations from Marx on the subject, I would like to make a few
preliminary points about the wider terms of reference of this debate. My
first point is that, despite Fred's best efforts, there does not appear to
me to be an "overwhelming" case to be made for the standard specification
of what I have called "socially necesary unproductive labor" (SNUL) as a
non-profit component of surplus-value. Marx's comments on how to handle
SNUL costs are neither systematic
no free of ambiguity. Fred might well be right that Marx entertained the
notion that currently produced surplus-value is divisible into two
components: profits and unproductive costs -- a notion that I have always
found problematic. But it seems to me that there are many passages in Marx
which suggest that he also entertained the possibility that these costs
should be subsumed under an expanded notion of constant capital. I use the
word "entertained" here advisedly, for my strongest impression is that
Marx never resolved the problem of how unproductive labor costs should be
treated in value-theoretic terms. This is an opinion that I have stated
openly on many occasions, including in my two major journal articles on
the subject (one of which was reprinted in my book Invisible Leviathan).
When I say he didn't "resolve" the problem, I mean two things. First, that
whether or not he ever arrived at a resolution of the issue in his own
mind, he failed to provide a written record of his position on the
question that is free of ambiguity; and second, that the written record of
his attempts to wrestle with the issue are unsatisfying in three major
respects: a) his failure to clarify fully the meaning of the phrase
"deduction from surplus value," b) his failure to explain his sometime
references to the capital exchanged with circulation labor as "variable
capital," and c) his failure to explore more fully the possibility that
SNUL costs should be treated as a special component of constant capital.

My second preliminary point is that Fred's "standard" specification of
SNUL as a component of currently-produced surplus value (i.e. as an
absolute "deduction" from aggregate surplus value) lends itself to the
view that SNUL constitutes a form of "luxury" expenditure on the part of
the social capital as a whole. This view is closely associated with
underconsumptionist tropes in debates over Marx's crisis theory. It
suggests that the investment that the social capital makes in SNUL -- in
order to sustain/increase the "annual rate of surplus value," to maintain
the capitalist state, and so on -- is a form of discretionary spending
that, in one period, can be increased to prime "effective demand," and, in
another period, decreased in order to expand the "profit component" of
social surplus value. This view of the matter is a bit too simple for my
taste. In the first place, it understates the "necessary" functions of
SNUL -- a point that is forcefully (and correctly) made by those who would
assimilate it to variable capital and who dispute the usefulness of the
productive-unproductive labor distinction; and in the second place it
overestimates the ease with which value can be reallocated from SNUL
activity to what Fred understands as the "profit component" of surplus

My third preliminary point is that the specification of SNUL as a
component of constant capital is a way of affirming that SNUL is indeed a
critically important component of the investment that the social capital
makes in the overall process of capitalist production and reproduction.
The capital invested in SNUL is not, of course, a component of the
constant capital STOCK, but it does form a stream within the FLOW of
circulating constant capital that finds expression in the total value of
the gross product. Moreover, SNUL plays a contradictory role with respect
to the vicissitudes of the average rate of profit: its socially necessary
functions help to sustain the latter (by, amongst other things,
counter-acting unfavorable conditions of realization that can "negate"
values), while at the same time its expansion involves an increase in the
overall investment that is being made to produce and realize surplus

This leads me to my final preliminary point. In empirical research, the
conceptualizing of SNUL as part of the constant capital flow allows us to
effectively remove SNUL costs from the direct calculation of the
fundamental Marxian ratios, since the constant capital FLOW, as distinct
from the constant capital STOCK, plays no role in these ratios. Hence:

s/C -- rate of profit, where the surplus value flow is calculated
primarily as a profit magnitude and C is the value of the constant
capital stock (which does not include SNUL);

s/v -- rate of surplus value, where v is the wage-bill of the productive

C/s+v -- organic composition of capital, conceived as a ratio of a stock
to two flows of new (current) value.

As it happens, my own research on the Canadian economy, conducted on the
basis of these definitions, has yielded results that are broadly
consistent with Marx's expectations regarding the long terms dynamics of
capital accumulation and the tendency of the average rate of profit to
fall. Moreover the trends in the Marxian ratios defined in the above
fashion seem to make sense, not only in terms of Marx's theoretical
expectations, but also in terms of what we know about the "real history"
of the Canadian political economy from 1947 to 1991 (see in particular my
1996 article in Studies in Political Economy).

The above considerations should go some way to making it clear why I
remain committed to the specification of SNUL as a component of the
constant capital flow rather than as a component of the flow of
currently-produced surplus value. So, with these considerations in mind, I
will now proceed to comment on a number of the points raised by Fred in
his 3 February post.

1. Fred states that "Murray seems to agree that the passages I had
presented in my last post do suggest that the costs of unproductive labor
are 'deductions from surplus value' and that the word 'deduction' implies
a simple subtraction, i.e. an absolute deduction and not a 'relative
deduction' as Murray has argued previously. Therefore, these passages
support the 'standard' interpretation..."

In fact, in my last post, I continued to defend the proposition that a
deduction from surplus value is best understood as a diminution in the
proportion of the total value of the commodity represented by surplus
value; for this is the only way that I can reconcile Marx's use of this
phrase with passages where he entertains a constant-capital specification
of circulation costs. I did concede, however, that I'm not happy with the
word "deduction" -- since it does indeed imply a simple subtraction
(or "taking away"). But the word "deduction" in this context is, to my
mind, still highly ambiguous, and its remains possible to entertain a
"relational" definition of its such that it refers to a "relative
diminution" rather than an absolute subtraction.

2. Fred continues:

"Murray argues that the only place where Marx clearly [?-MS] said what he
meant by 'deduction from surplus value' is from the Grundrisse (p. 548) in
which it is described as an 'increase of necessary labor in relation to
surplus labor'. Murray continues: 'This definition suggests that, as a
proportion of the value of the commodity, surplus value will shrink as a
result of an increase in the value of the elements of cost-price (constant
capital or variable capital or both).' But I do not see how Murray's
conclusion follows. Murray seems to be defining necessary labor to
correspond to both constant capital and variable capital. However, the
concept of necessary labor, as I understand it, is not related to constant
capital in any way."

But this is precisely the point at issue: does socially-necessary
unproductive labor form a special component of constant capital? Is it
unreasonable to call the labor involved in circulation "necessary labor"
even though it is not also creative of surplus value? It seems to me that
Marx's (not-so-"clear") definition of a "deduction from surplus value" as
"an increase of necessary labor in relation to surplus labor" actually
favors the constant capital interpretation of SNUL precisely because it
suggests that circulation labor costs represent an "increase of necessary
labor." If we bring together the contested quotes from Capital II and the
Grundrisse, we arrive at the following series of statements:

1) circulation labor costs = deduction from surplus value;

2) deduction from s.v. = increase in necessary labor


3) circulation labor costs = increase in necesary labor;

and since,

4) circulation labor costs involve unproductive labor;


5) unproductive labor in circulation = an increase in necessary

This would seem to support the idea that Marx saw some types of
unproductive labor (and circulation labor in particular) as a form of
"necessary labor." So far, it seems to me, the contested passages lend
support to the Mage-Smith category of "SNUL."

Fred seems to be saying, against this interpretation, that:

1) deduction from s.v. = non-profit component of s.v.;

such that,

2) non-profit component of s.v. = increase in necessary labor
in relation to surplus labor;

but since "non-profit" surplus-value is still a monetized form
of surplus labor, then:

3) a component of surplus labor = an increase in necessary labor
in relation to surplus labor.

I think it should be fairly clear why I see an elementary logical problem
with an interpretive reading of Marx that yields proposition (3) as its

Fred continues:
"Necessary labor is part of the total labor of the CURRENT period of

I agree, so long as it is understood that this necessary labor contains
two very different components: socially necessary productive labor (SNPL)
and socially necesary unproductive labor (SNUL).

Fred continues:
"The total labor of the current period produces a certain quantity of
value added (what Marx often called 'new value'). This quantity of 'new
value' is then divided into two parts: one which is necessary to replace
the variable capital invested and the remainder which is surplus-value."

Apart from the fact that I have trouble with the notion of "variable
capital invested," I find this unobjectionable.

Then Fred says:
"Necessary labor is the number of current hours required to produce 'new
value' equal to the variable capital. Surplus labor is the rest of the
working day."

Here we must part company, for now Fred is equating necessary labor with
SNPL, which is indeed the only component of "necessary labor" that
PRODUCES new value. By definitional fiat, Fred is excluding the category
of SNUL which is implied in the passages from Marx cited above.

Fred continues:
"Thus, necessary labor has nothing to do with constant capital or the
recovery of constant capital. Necessary labor is part of CURRENT labor;
constant capital, on the other hand, is not replaced by current labor, but
rather from previously existing value which is transferred to the price of
the product."

Obviously, I disagree with the first sentence. In my view, however, there
is no problem with affirming that SNUL is part of current labor AND
simultaneously affirming that it is a component of constant capital.
Constant capital -- that capital which makes an indirect contribution to
valorization -- has many phenomenal forms in the overall process of
production/reproduction: from machines of production, to machines of
realization, to machines of institutional reproduction. Just as an
industrial robot is a part of CURRENT production, an element of constant
capital that is "combined" with necessary productive labor in producing a
new product, SNUL is involved in CURRENT processes of circulation,
realization, state maintenance, and so on, and as such is a component of
CURRENT NECESSARY LABOR. But unlike SNPL, SNUL does not create the value
that it is exchanged with -- any more than a machine used in production
creates the value with which it is exchanged. The elements of constant
capital -- a punch press, a ton of sheet metal, or the wages of a sale
clerk -- represent value, transfer value, and re-present value. Their
existence is predicated on the past creation of surplus value and its
accumulation as capital. Thus, of course, "constant capital is not
replaced by current labor, but rather from previously existing value which
is transferred to the price of the product."

Reminding us of what Marx said about necessary labor in Capital I (where
Marx abstracts from the costs of circulation entirely), Fred concludes
this line of his argument with the statement that "Necessary labor has
nothing to do with constant capital." Well, we know that's your position,
Fred, but this is more a definitional assertion than a conclusive
argument. I would agree with you that SNPL "has nothing to do with
constant capital" -- except of course in the sense that it is a factor
in "conserving" its value. But it doesn't follow from this that SNUL "has
nothing to do with constant capital."

Fred then returns to the Grundrisse passage where Marx states that a
deduction from surplus value is an increase in necessary labor in relation
to surplus labor, and states that:
"Necessary labor is here defined as the PRODUCTIVE [my emphasis -- MS]
labor time necessary to reproduce an equivalent both of the variable
capital and the costs of circulation labor, and surplus labor is the
difference between the total productive labor time and this necessary
labor, as just redefined."

I disagree with the "definition" that Fred is attributing to Marx here.
Marx's reference to an "increase" in necessary labor is clearly a
reference to the additional necessary labor represented by circulation
labor, and, as we all know, Marx did NOT regard circulation labor as
productive. It is a legerdemain for Fred to suggest that Marx is defining
all of the necessary labor in this context as "productive" -- such that
the expenditure of ciculation labor does not form a component of
"necessary labor" but requires additional productive labor to "produce an
equivalent" of its cost. In the end, Fred is simply ruling out, by
definitional decree once again, the very admissibility of the category of
SNUL as a component of "necessary labor."

Fred continues:
However, this does not mean that necessary labor also includes the time
necessary to reproduce the equivalent of the constant capital. Indeed,
this would contradict Marx's basic theory of value. Necessary labor is a
part of current labor and constant capital is not recovered from the value
produced by current labor."

Again, Fred's refusal to distinguish SNPL and SNUL muddies the issue in
dispute. It is true that SNPL does not include "the time necessary to
reproduce the equivalent of the constant capital." But SNUL is a "part of
current labor," AND, as one component of constant capital, its value (i.e.
the value represented by its wage bill) is not recovered from the value
produced by current SNPL (SNUL, itself, of course, being incapable of
producing value).

Fred sums up this part of his argument in the following way:
"...this passage about necessary labor in the Grundrisse reinforces the
standard interpretation, instead of contradicting it. If these costs are
recovered out of the value added by current labor, then they cannot be a
part of constant capital."

But this conclusion follows ONLY if one accepts Fred's own definitions. On
my definition of the concepts in dispute, I can easily assert (without
thereby proving, of course) the following counter-proposition:
"If these costs represent the previously-existing value added by the wages
of currently-employed SNUL, then they are a part of constant capital."

Fred continues that Marx "described" what he meant by a "deduction from
surplus value" not only in the contested Grundrisse passage, but also in
the famous Capital II passage which reads:
"The replacement of these [circulation] costs must come from the surplus
product, and from the standpoint of the capitalist class as a whole it
forms a deduction of surplus-value or surplus product..."
Fred is suggesting that Marx is defining "a deduction from surplus value"
not only as an "increase in necessary labor in relation to suplus labor"
but also as a "replacement of [circulation] costs [that comes] from the
surplus product." Okay, let's try this out. Substituting the Grundrisse
definition for the "deduction" phrase in Capital II, we get:

1) "The replacement of these costs must come from the surplus
product, and from the standpoint of the capitalist class as
a whole it forms an increase in necessary labor in relation
to surplus labor."

Now, since Fred is insisting on defining "necessary labor" here as current
productive labor, we arrive through a further substiution at the

2) "The replacement of these costs must come from the surplus
product, and from the standpoint of the capitalist class as
a whole it forms an increase in necessary productive labor in
relation to surplus labor."

This is a very odd statement to say the least -- and certainly far from
unambiguous. What does it mean to say that "the replacement of circulation
costs" FORMS an increase in necessary PRODUCTIVE labor? If Marx intended
to say what Fred attributes o him, why didn't he say that these costs
REQUIRE an additional outlay of productive labor? For that matter, if Marx
was arguing that SNUL should be treated as a non-profit component of
surplus value, why didn't he say that in so many words? Why didn't he
refer to these circulation costs as an ALLOCATION of surplus value instead
of as a DEDUCTION of "surplus value or surplus product"? And why the
equivocation implicit in the expression "surplus value or surplus
product"? Was Marx in his own mind distinguishing between and equivocating
over two possibilities -- on the one hand, that these costs are
"recovered" out of currently-produced surplus value and, alternatively,
that they are "covered" by value represented in the
historically-constituted surplus product (i.e. the previously existing
value represented by the various elements of constant capital)?

A great deal seems to hang on the interpretation of a single passage whose
meaning is far from transparent, and which in any case forms a part of a
manuscript that Marx never completed for publication. I can't share Fred's
confidence that his interpretation is the only reasonable one -- and I
certainly don't see how this "general law" passage in Capital II can be
used to invalidate the assimilation of SNUL to constant capital.

3. Fred expresses the opinion that "Marx's 'general law' means exactly
what it says -- that no value is produced in exchange and that circulation
costs must therefore be recovered out of surplus-value." I certainly agree
that "no value is produced in exchange," but I don't see how it follows
that circulation costs are necessarily recovered out of currently-produced
surplus value. As I argued in my previous post, it's necessary to
distinguish between the addition of previously-existing value to the
commodity in the sphere of circulation and the addition of new value. To
say that old value can be added to the commodity by circulation costs is
not to imply that circulation labor produces value. My view is that the
costs of circulation -- and other SNUL expenses -- are attached to the
constant capital component of the value of the commodity. It is not
"exchange" that effects the transfer of these previously-existing values
to the commodity product, but the process of capitalist reproduction taken
as a whole.

4. Fred turns next to the "final passage left to support Murray's
argument," the Capital III passage (pp. 405-406) which includes the
statement: '[The] additional value that [the merchant] adds to commodities
by his expenses is reducible to the ADDITION OF PREVIOUSLY EXISTING
VALUES, even though the question still arises here as to how he
maintains and conserves the value of this CONSTANT CAPITAL" (emphasis
added). On my interpretation, the costs of circulation (including
commercial SNUL) are here referred to as additions to the value of the
commodity, as an "addition of previously existing values," and as
"constant capital." Preparing the ground for his counter-interpretation,
Fred makes much of the fact that this parenthetical remark is to be found
in a sentence which includes the phrase: "...the merchant, being simply an
agent of circulation, produces neither value nor surplus-value...."
Indeed Fred suggests that the latter phrase "appears to be contradicted"
by the parenthetical remark about the "addition" of previously existing
values. The only way out of this contradiction, he implies, is to
interpret the "additional value" in a way that is different from what Marx
seems to intend. Before proceeding to a consideration of Fred's
reinterpretation, however, it should be noted again that there is NO
CONTRADICTION between affirming that an agent of circulation cannot
PRODUCE value or surplus-value and affirming that that same agent's
costs/wages contributes ("ADDS") to the value of the commodity
previously-existing, constant-capital values equivalent to costs incurred.
It's unclear to me why Fred insists upon equating "adding value" with
"producing value." After all constant capital in the sphere of production
"adds value" without "producing value" (for only living productive labor
"produces" or "creates" value).

In any case, Fred then proceeds to suggest that Marx's reference to
"previously existing values" in this context refers not to "constant
capital" (even though Marx himself makes explicit reference to it as
constant capital), but to surplus value unrealized by productive capital
when it sold the commodity at a "wholesale" price to the merchant. Thus,
when the merchant sells the commodity at the retail price, the merchant is
able not only to realize a commercial profit, but also "previously
existing values" that may be used to recoup business expenses. The
"previously existing values" are thus really a part of the "currently
produced surplus value" that industrial capital shares with commercial

I'm sure that Fred will not be surprised when I tell him that I find this
interpretation as implausible as he seems to find my interpretation of
Marx's reference to a "deduction from surplus value or surplus product."
In the first place, Fred's interpretation, ingenious though it is, fails
to account for why Marx would refer to the "ADDITIONAL value that [the
merchant] ADDS to commodities by his expenses" if what he really meant was
that "previously existing current surplus value" has been transferred to
the merchant from the producer so that the merchant can recoup his/her
costs as well as realize a profit. In the second place, it doesn't explain
why Marx associates "previously existing values" explicitly with constant

Fred seems to be on stronger ground when he quotes another passage from
Capital III where Marx states that:

"In the supplementary equalization of profits brought about by the
intervention of commercial capital, we saw that no additional element goes
into the value of the commodity for the money capital that the merchant
advances, and that the addition to the price, whereby the merchant makes
his profit, is simply the portion of commodity value that productive
capital has not included in the production price of the commodity, and has
in fact left out." (p.401)

In the second part of this statement, Marx is saying that commercial
profit is a portion of surplus-value that productive capital does not
realize when the commodity is sold to commercial capital, and which is
only realized by the merchant following the sale of the commodity to its
final buyer. He does NOT state that the portion of the commodity value
that productive capital has not included in the production price of the
commodity also covers "the money capital that the merchant advances."
In the first part of the sentence, I therefore understand Marx to be
saying simply that the merchant's capital makes no contribution to the
CREATION of value or surplus-value, even though it does participate in the
redistribution of surplus value on the basis of the general rate of profit
(formed by "the supplementary equalization of profits brought about by the
intervention of commercial capital"). It is of course a commonplace that
commercial profit is surplus value transferred from productive capital to
commercial capital in return for commercial capital's important service in
trying to realize the full value of the commodity in the market.

Fred seems to interpret the above-quoted passage in a way that would make
it difficult to reconcile with another Capital III passage (one referred
to by Rakesh on 3 February) which lends some support to a constant-capital
specification of SNUL. The passage states:

"The expenditure [on commercial workers], even though incurred in the form
of wages, is distinct from the variable capital laid out on the purchase
of productive labour. It increases the outlays of the industrial
capitalist, the mass of labour he has to advance, without directly
increasing surplus-value. For this is an outlay for labour employed simply
in realizing values already created. Just like other outlays of the same
kind, this too reduces the rate of profit, because the capital advanced
grows, but not the surplus-value." (p.413)

In principle, I see no reason to treat the capital advanced by commercial
capital for commercial workers' wages any differently. Just as the capital
of the industrial capitalist must be "made good" out of the value
objectified in commodities, so too must the advanced capital of the
merchant. Marx writes in a related vein in Capital III (Progress edition,
1978, pp. 291-292, which presents a clearer translation of the passage
found on p. 405 of the Vintage edition):

"These [circulation costs] form additional capital, but do not create
surplus value. They must be made good out of the value of the commodities,
because a portion of the value of these commodities must be reconverted
into these circulation costs. But no additional surplus value is created
thereby. So far as this concerns the total capacity of society, it means
in fact that a portion of it must be set aside for secondary operations
which are no part of the self-expansion process, and that this portion of
the social capital must be continually reproduced for this purpose."

Here Marx assimilates the circulation costs of industrial capital with
that "portion of the social capital [that] must be continually
reproduced" -- after the fashion of constant capital in production. (Note
that he does NOT speak of "recovering" these costs out of current surplus
value.) There are other passages that strongly suggest that Marx was
"entertaining" a constant capital specification of SNUL, passages that I
have reproduced in my journal articles as well as in Invisible Leviathan.
All of this, I think, casts considerable doubt on Fred's claim that the
"textual evidence overwhelmingly supports the standard interpretation."

I've gone on for too long, and I must bring this to a close. I want to do
so by paying tribute to Fred's work on Marxist crisis theory and the
falling rate of profit, which, in some respects, is quite close to my own.
Operationally, we are not all that far apart in the way that we go about
"testing" Marx's hypotheses. Having said that, I'm quite convinced that
our differences will not be settled by appeals to this or that passage
from Marx. There seems to be room for more than one interpretation of how
Marx would specify SNUL in relation to his value categories. But the proof
of the pudding is ultimately in the eating. In the last analysis, the
reason that I find the constant-capital specification of SNUL so
compelling is that it makes good intuitive sense to me, it avoids the
fetishistic reduction of constant capital to "physical means of
production," it meets many of the objections that have been made by
critics of the productive-unproductive labor distinction, and it yields
damned interesting results -- results that seem to accord with visible
economic trends, confirm Marx's theoretical prognostications, and resist
appropriation by reformist political projects. I think that those are all
pretty good reasons to favorably entertain the "Mage heresy" -- reasons,
I think, that Marx himself would have approved.